Genetic selection for feed efficiency (FE) could decrease feed costs on dairy farms. Currently, measuring individual feed intake on commercial dairies is impractical due to high equipment and labor costs. Development of portable, affordable technologies as indicator traits would help make selection for FE feasible. The objective of this study was to determine if traits collected by automated sensor technologies were associated with feed intake. Furthermore, the impact of an animal’s health status on proxies was examined. Data were collected on 108 lactating Holstein cows (parity 1–4, DIM 50–215) in 2 seasonal groups; summer (n = 48) and fall (n = 60). All animals were fit with eartags tracking activity (ETACT) and inner ear temperature (ETTEMP). Fall cows also received an additional eartag measuring rumination and rumen boluses recording rumen pH (RBpH), temperature (RBTEMP) and cow activity (RBACT). Individual feed intake, milk weights, components, body weight (BW), BCS, and health events were recorded. Data were analyzed with PROC GLIMMIX in SAS. Models included dry matter intake (DMI) as the response variable and fixed effects of DIM, milk weight, fat, protein, and lactose, metabolic BW, sensor measure, and health event (HE). Random effects included parity, pen and seasonal group. When animals were classified by their HE(s) for the entirety of the trial, RBpH and RBTEMP were associated with DMI (P < 0.01), while RBACT tended toward significance (P < 0.1). Each interaction between HE and RBpH, RBTEMP and RBACT were also significant (P < 0.0001). When mastitis was included at the time of illness, ETTEMP, RBpH and RBACT were associated with DMI (P < 0.05). When lameness was examined at the time of the event, ETTEMP and RBACT were significant (P < 0.05), while RBpH tended toward significance (P < 0.1) Results indicate automated sensor traits act as indicators of feed intake. Health events appear to have long lasting influence on sensor trait and feed intake phenotypes. Future work will focus on development of models to predict feed intake from these data.